Aubot LMS Content Automation
Improving curriculum delivery efficiency through low-code automation
(Google Sheets + App Script)
Aubot: Education website teach children computer science
Your role: Software specialist (Developer + UX designer + Business analyst)
Automation process: 1 month
Background: At the time, the platform structure was already in place. My role, alongside a small dev team, was to implement educational content—turning detailed curriculum documents into interactive lessons—within the LMS.
We were using Postman to manually call backend APIs and update the database with each lesson and exercise.
Tasks: Convert over 20 chapters of Python content into structured data entries for the LMS, including multiple types of exercises such as:
Multiple-choice
Code input
Fill-in-the-blank
Short answer
Each chapter required careful formatting and testing to align with backend data requirements.
Proposal
Proposal
1. Assess the Workflow (Process Analysis & Documentation)
Methods used:
Process mapping (using diagrams and bullet steps)
Task timing (I measured average time per lesson: 20–30 mins)
Error sampling (noted 3–5 typical issues during manual API calls)
2. Interview Stakeholders (Informal Research & Validation)
Methods used:
One-on-one interviews
Note-taking and keyword clustering (grouping pain points)
Input/output tracking for typical document-to-JSON conversion
Pain Points
Despite having a functional backend and API, the content implementation workflow was not scalable:
High dependency on technical team
All content had to be entered manually by developers using Postman.
→ This created a bottleneck as only technical staff could contribute to progress.Inefficient use of time and resources
Each lesson update involved copying JSON, formatting it, triggering API calls, and checking responses manually.
→ The process was slow, error-prone, and inconsistent across team members.No standardized input structure
Curriculum documents varied in structure and were not ready for backend use.
→ Required constant back-and-forth between content creators and developers.Lack of collaboration tools for non-technical team members
Non-developers couldn’t participate in implementation or testing.
→ Reduced transparency and team velocity.Tight delivery timeline
With 4 months until user testing and over 20 chapters to implement, continuing the manual approach would risk missing deadlines and lowering content quality.
Solution Design
Based on the pain points and stakeholder needs, I proposed a low-code automation system using Google Sheets and Google App Script.
Before finalizing this solution, we researched and compared several tools and approaches — including writing a standalone Python script, using Airtable automations, and exploring integration platforms like Zapier.
Requirement
Avoid developer bottlenecks
Reduce formatting errors
Simplify API calls
Enable version control
Original workflow
Open Postman and the content documents
Optimised workflow
Design Choice
Allow content creators to contribute via spreadsheet
Pre-define column structures & dropdown validations
Use App Script's UrlFetchApp() for 1-click submission
One sheet per chapter, allowing collaborative editing
Insert content into JSON format via Postman
Repetitive work and require manual checks
Identify where to place fields such as title, options, and correctAnswer
Escape special characters (e.g., convert " to \") to avoid breaking the JSON structure
Double-check formatting manually
Open content document
Select and copy the exercises
Copy exercise content
Paste into Google Sheet
Submit the JSON to backend API via Postman
Submit via button (App Script trigger)
Implementation
My Key Contributions:
1. Functional Testing
Tested the script across different content formats (multiple-choice, fill-in-the-blanks, code-based questions)
Verified JSON structure and API responses using staging environment
Logged and reported issues such as formatting inconsistencies, unexpected errors, and input edge cases
2. User Testing & Feedback
Observed how non-technical team members interacted with the tool
Documented usability friction and recommended improvements (e.g. clearer success/failure messages, better field instructions)
3. Bug Reporting & Iteration
Created a shared issue tracker to log bugs and observations during testing
Worked closely with the developer to replicate and fix issues
Re-tested after each iteration to confirm resolution
4. Internal Enablement
Created a quick-start guide and documentation for the team
Walked through the tool in internal sessions to onboard content creators
Answered implementation-related questions and helped team members adopt the new workflow
Impact & Results
Improved internal documentation & process transparency
Created guides, templates, and trackers to support onboarding and long-term maintenance
Documentation made future scaling and handover easier
Scalable foundation for future content growth
The framework supports other subjects (e.g., Java, Robotics) and is already being reused
Potential for further integration with AI-based curriculum tools and auto-documentation
Implementation speed increased by over 5×
Manual implementation of one chapter previously took 3+ hours
With the new workflow, the same task could be completed in under 30 minutes
This enabled us to finish the full Python curriculum in under 3 months, one month ahead of schedule
Non-technical team members were empowered to contribute
Content creators, previously dependent on developers, could now independently implement lessons with minimal training
Reduced internal bottlenecks and increased parallel workflows
Reflection & Future Potential
This project taught me how even simple, low-code solutions can drive meaningful impact when grounded in real workflow insights and stakeholder needs.
Looking ahead, this automation framework could be extended with version control, integrated documentation, or even AI-powered lesson generation — making it an even more powerful tool for scalable curriculum development.